A large number of patients have been treated with murine monoclonal antibodies in several phase I and II studies. In general, the anti-tumor effects of this therapy have been disappointingly few. It has been demonstrated, however, that antibodies can reach the site of the tumor and strain the tumor cells specifically. In general it appears that monoclonal antibodies in and of themselves as currently being used are not effective tools in the treatment of cancer. One limiting feature of the use of murine monoclonal antibodies has been the development of human antimouse antibodies. If antibodies ultimately are modified to become an effective antitumor treatment for patients with cancer, the development of human antimouse monoclonal antibodies will still be a significant problem. We, therefore, initiated a study using human monoclonal antibodies in an attempt to treat patients with colorectal carcinoma. Two human monoclonal antibodies 16.88 and 28A32 are currently being studied. Both are derived from human B cells from patients with colorectal carcinoma immunized with an autologous tumor cell vaccine. These monoclonal antibodies have been radiolabeled with I-131 and have localized to tumors in patients with metastatic colorectal carcinoma. Of 14 patients treated so far, none have developed antibodies against the administered human monoclonal antibody. However, two patients had pre-existing antibody directed against either or both monoclonal antibodies. These patients were not entered on the study. Despite trafficking of these monoclonal antibodies to the tumor-bearing site, no antitumor responses have been observed.